Representation Without Consultation: Three Cases the Literature Collapses

may 24, 2026

A real question in negotiation ethics: what do you owe a third party whose interests are on the table when you can’t talk to them directly? Stakeholder theory says workers count as stakeholders. Disability rights organizing gave us “nothing about us without us.” Both frames are sharper than the practice that invokes them, and the gap is where the actual work lives.

The frames as they’re usually deployed

Stakeholder theory’s contribution was to widen the field of who an organization owes consideration to. Workers, communities, suppliers, sometimes the public: all of them affected by a firm’s decisions, all of them with standing in the moral accounting. The vocabulary did real work in moving corporate ethics past pure shareholder-primacy. Where it gets thin is at the operational layer. Saying a group has standing does not specify how that standing gets exercised, and stakeholder theorists have been criticized (Charles Blattberg’s the standard cite) for treating negotiation between fixed interests as the privileged mechanism, when the more demanding question is how interests get formed and revised in the first place.

“Nothing about us without us” is the more aggressive frame. Originating in sixteenth-century Polish parliamentary politics (the Nihil novi constitution), revived in the 1990s by disability rights activists in South Africa, codified into the Convention on the Rights of Persons with Disabilities, the principle is procedural: no decision affecting a group can be made without that group’s participation. It is stronger than stakeholder theory because it specifies the mechanism (participation, not just consideration), and weaker in that it gives less guidance about what to do when participation is genuinely infeasible.

Marcel Paret’s work on South African unions adds the negative case. Even movements explicitly built to serve workers can slip into paternalism: organizers substituting their own judgment for the workers they organize, educating-toward rather than organizing-with, disciplining grassroots energy that doesn’t fit the strategy. Paret’s interviews with COSATU activists during a 2014 protest show this happening in real time among people who would identify themselves as the opposite of paternalists. The slip is structural, not intentional.

Each of these frames is doing genuine work. Together they push toward a default: if you are about to act on behalf of a third party, your first move is to find a way to involve them. Anything less risks the failure modes the literature has named.

Where the frames stop being adequate

The cases the frames handle cleanly are cases where consultation is available. You could ask the workers; you chose not to. You could include the disabled community in the policy design; you opted for expert proxies. Those are the clean cases, and the frames give the right answer: don’t do that.

The harder cases, the ones I think the literature mostly collapses, are the ones where consultation isn’t a simple option. At least three structurally different situations get treated as the same case:

(a) You can consult but choose not to. This is the situation the literature is mostly responding to. The disability lit grew up in opposition to expert-driven policy made over the heads of disabled people who were available to be asked. Paret’s union activists could have asked the unemployed what they wanted; they spoke for them instead. The diagnosis is correct here: it’s paternalism, and the prescription (consult) is right.

(b) You can’t consult easily. A negotiator may not have a channel to the affected staff. A non-profit advocating for a class of workers may have no roster, no mailing list, no way to convene the people whose conditions are being negotiated. This isn’t paternalism by choice; it’s representation under structural constraint. The “nothing about us without us” principle, applied naively, says: don’t act. But not-acting also has costs to the group, often higher costs than acting imperfectly on their behalf.

(c) Consulting would itself harm them. A staff member whose employer is hostile to outside advocacy may be exposed to retaliation if known to be consulted. Disclosing the negotiation prematurely could compromise leverage that ultimately serves the group. This is the case where the procedural principle and the substantive one come apart. Consultation, the very mechanism that “nothing about us without us” exists to protect, becomes the thing that endangers the people it is meant to protect.

The literature mostly addresses (a). It has a lot less to say about (b) and (c), and the field-experience evidence suggests that practitioners faced with (b) or (c) tend to fall into one of two failure modes: (i) treating the case as (a) and refusing to act, or (ii) treating the case as (a) and acting anyway, while invoking constraint as a free pass that exempts the action from the procedural critique.

Neither is right. The honest position is that (b) and (c) require different ethics than (a), and the literature hasn’t done that work in a way most practitioners can find.

What the ethics actually requires

I think the substantive principle is something like this:

For (b), the test is not “did you consult” but what did you do about the inability to consult. Did you treat the constraint as permanent and build a representational structure around it (paternalism by inertia), or did you treat the constraint as a problem to solve and work to create the conditions for direct voice over time (organizing in the older sense of the word)? Paret’s actual contribution, on a careful read, is in this direction. Union paternalism becomes the problem when it stops being a bridge and becomes the structure. Acting on behalf of people who can’t currently be reached is defensible if your endgame includes reaching them.

For (c), the test is whether the protective justification is real or rhetorical. Real protective representation has the property that consultation, when eventually possible, ratifies the choices that were made on the group’s behalf. Rhetorical protective representation has the property that the represented group, once consulted, says something other than what the representative claimed. The disability community has lived this distinction for decades. There is a long history of well-meaning advocates discovering that the population they spoke for wanted something quite different, and the experience has produced a strong norm that “I knew what they really needed” is the most dangerous sentence in advocacy. The ethics requires that you build in mechanisms for the represented group to override your representation when it becomes possible for them to do so, and that you treat any override as data, not as ingratitude.

The bridge concept is the link between (b) and (c) and the cleaner cases. Representation that points at its own dissolution, that aims to make itself unnecessary, is one thing. Representation that perpetuates the conditions of its own necessity is something else. Most institutional advocacy drifts toward the second because the second is what funds the institutions; this drift is structural and is the thing organizers who think about this for a living spend most of their energy trying to resist.

What the frames could absorb

Stakeholder theory could be sharpened by distinguishing co-stakeholder relationships (the affected party participates) from beneficiary relationships (the affected party is acted upon for their good). The terminology is in the literature but is not deployed in the negotiation contexts where it would do the most work. A negotiator who is clear about whether the staff of an organization are co-stakeholders or beneficiaries in a given negotiation will make different operational choices, and the choices that get defended as “stakeholder consideration” in the corporate context are very often beneficiary-shaped choices wearing co-stakeholder language.

“Nothing about us without us” could be sharpened by acknowledging that the principle is aspirational with respect to mechanism, not deontic with respect to outcome. The aspiration is participation. When participation is infeasible, the principle becomes: act in a way that makes participation more feasible going forward, and treat eventual participation as the test of whether your representation was honest. This is more demanding than the principle as commonly invoked, not less.

And the union literature, Paret most directly, could foreground that paternalism is a process, not a position. Activists do not declare themselves paternalists. They become paternalists through the accumulation of small substitutions that each looked, in the moment, like the only way to act. The remedy is not a vow but a practice: routine audit of where the represented group has been substituted-for, with the bias set toward concluding too quickly that substitution has happened rather than too slowly.

What this resolves

It does not resolve cases. Cases stay hard. What it does is offer a sharper diagnosis: when you find yourself negotiating on behalf of a third party you have not consulted, the question is not “am I a paternalist,” because the answer when stated that flatly is always no. The question is which of the three cases you are in, what your endgame for the represented group’s voice is, and whether you would still defend your action if the group, once able to speak, said something different.

That last test is the one that matters. Everything else is method.

if it stayed with you, write to me.